I am currently working on the detection of virus from bacterial metagenomes. I have a blastp file obtained from a blastp analysis using refseq viral protein database. When I import this file to MEGAN with a min score =50 and evalue = 1e-05, min support=10, top percent=10, I obtain a high proportion of reads in the not assigned category. When inspecting these reads, I can observe that most of them show a significant identity with hits of virus, whose scores are between 40-49. In this case, I would like to know if changing min score parameter =40 is recommendable in order to to reassign all these reads?